Estimation of bias and variability of continuous gas monitor data: application to carbon monoxide monitor accuracy.
Shulman SA; Smith JP
Cincinnati, OH: U.S. Department of Health and Human Services, Public Health Service, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health, EPHB 005-128, 2001 Jul; :1-58
A method is presented for the evaluation of the bias, variability, and accuracy of gas monitors. This method is based on using the parameters for the fitted response curves of the monitors. Thereby, variability between calibrations, between dates within each calibration period, and between different units can be evaluated at several different standard concentrations. By combining variability information with bias information, accuracy can be assessed. An example using carbon monoxide monitor data is provided. Since monitor response can be determined in a chamber simultaneously for several monitors, correlation of errors about the fitted lines is considered. Monitor drift between calibrations is also considered. Procedures for variance components estimation are provided for balanced data. Procedures using Proc Mixed in SAS that provide solutions for both balanced and unbalanced data are available from the authors.
Region-5; Monitoring-systems; Monitors; Data-processing; Measurement-equipment
Field Studies; Control Technology
NTIS Accession No.
National Institute for Occupational Safety and Health